1 Insights Into Transporter Classifications: an Outline Of

Home , ABCC8, ABCC9, ABCE1, ABCF1, ABCF2, Antiporter

1 Insights into Transporter Classiﬁcations: an Outline of Transporters as Drug Targets Michael Viereck,1 Anna Gaulton,2 and Daniela Digles1

1University of Vienna, Department of Pharmaceutical Chemistry, Althanstrasse 14, 1090 Vienna, Austria 2EMBL – European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, CB10 1SD, UK

1.1 Introduction

Classifications are a useful tool to get an overview of a topic. They show instan- ces grouped together that share common properties according to the creator of the classification, and as in the case of hierarchical classifications, they allow to draw conclusions on the relation of different classes. In the case of the identifica- tion of drug targets (including transporters as the main drug target), several pub- lications show classifications as a helpful tool. Imming et al. [1] categorized drugs according to their targets, to get an estimate on the number of known drug targets, including channels and transporters. Drugs on the market were connected to a target only if it was described as the main target in the literature. For transport proteins (including uniporters, symporters, and antiporters), this identified six different types of transporter groups that are relevant as drug targets. These are the cation-chloride cotransporter + + + + (CCC) family (SLC 12), Na /H antiporters (SLC 9), proton pumps, Na /K ATPases, the eukaryotic (putative) sterol transporter (EST) family, and the neu- + rotransmitter/Na symporter (NSS) family (SLC 6). These families mainly belong to either ATPases or solute carriers (SLC). Whether the EST family is treated as transport protein or not depends on the classification used. Rask-Andersen et al. [2] used a manually curated and extended version of the DrugBank [3] data from 2009 to analyze drug targets. They identified 435 therapeutic effect-mediating targets, where the third largest group (67) is of transporter proteins (including 35 ion channels). These transport proteins are mainly targeted by antihypertensive drugs, diuretics, anaesthetics, and antiarrhythmic drugs. In a more recent study [4], they analyzed the Drugs in the Clinical Trials Database by CenterWatch to investigate the targets of new drug candidates.

Transporters as Drug Targets, First Edition. Edited by Gerhard F. Ecker, Rasmus P. Clausen, and Harald H. Sitte.  2017 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2017 by Wiley-VCH Verlag GmbH & Co. KGaA. 2 1 Insights into Transporter Classiﬁcations: an Outline of Transporters as Drug Targets

Transporter proteins in this data set were classified using the transporter classification (TC) system established by Saier et al. [5]. This chapter will first give an overview on selected existing transporter classifications and then describe our process of creating a combined classification scheme for the ChEMBL [6] database. Finally, the investigation of the counts of drugs and diseases for one example protein superfamily is provided, to show the usefulness of classifications in characterizing related proteins and to give a first overview on the topic of this book, focusing on transporters as drug targets.

1.2 Available Transporter Classiﬁcations

As a consequence of the significance of the transport process for every living organism, already numerous classification schemes for membrane transport proteins of several organisms exist. Quite a few focus only on specific families, for instance, the SLC superfamily or the ABC transporter superfamily. Table 1.1 shows a selection of membrane transport protein databases with a focus on human proteins. To keep the table concise, databases focusing on different organisms (e.g., the plant membrane protein database Aramemnon [7], the yeast transport protein database YTPdb [8], or ABCdb [9], a database about bacterial ABC systems) are not included, even though bacterial or protozoal channels and transporters can also be promising drug targets [10,11]. We were interested in classification schemes that not only try to cover the full variety of human membrane transport proteins but also provide their own complete classification. Therefore, we analyzed the functional and phylogenetic classification scheme of the Transporter Classification Database (TCDB), the more pharmacology-driven IUPHAR/BPS classification, and the mainly functional- driven classification of channels and transporters in the bioactivity database ChEMBL-16 [18]. These were recently reviewed in [19], which provides a more detailed discussion for the interested reader. In addition, we included the SLC series [12], which is a well-known nomenclature system.

1.2.1 TCDB

The transporter classification system developed in the laboratory of Saier is to some extent comparable to the Enzyme Commission (EC) classification. But while the EC system concentrates on the function of enzymes, the transporters in the TC system are classified according to function and phylogeny [20]. Both schemes are recommended by the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology [21,22] (URL: http://www.chem .qmul.ac.uk/iubmb/mtp/). The TC is also included in several other databases such as UniProt [23] (URL: www.UniProt.org/) or the Protein Data Bank (PDB) [24] (URL: http://www.rcsb.org/pdb/). TCDB is an exhaustive classification Table 1.1 Selection of transporter collections with a focus on human membrane transport proteins.

Database Description URL Included proteins Limited to organisms

Bioparadigms SLC Online resource of the 52 human slc.bioparadigms.org/ Solute carriers Homo sapiens Series [12] solute carrier families ChEMBL [6] Large-scale bioactivity database for https://www.ebi.ac.uk/chembldb/target/ Proteins with bio- — drug discovery browser activity data Human ABC Basic information about human www.nutrigene.4t.com/humanabc.htm ABC transporters Homo sapiens transporters [13] ABC transporters IUPHAR/BPS Guide to Overview of human drug targets www.guidetopharmacology.org Drug targets Homo sapiens, and PHARMACOLOGY [14] with their pharmacology Mus musculus, Rattus norvegicus TCDB [5] Provides a classiﬁcation scheme www.tcdb.org/ Channels, transport- — for all membrane transport pro- ers, auxiliary trans- teins in all living organism port proteins TransportDB [15] Genomic comparison of mem- www.membranetransport.org/index.html Channels, transport- Complete genome-

brane transport proteins, predic- ers, auxiliary trans- sequenced organisms Classi Transporter Available 1.2 tion of their function, and port proteins (365 total) classification according to TCDB Transporter substrate Provides a transporter substrate http://tsdb.cbi.pku.edu.cn/home.cgi Transporters — database (TSdb) [16] repository with mappings to KEGG pathways UCSF-FDA Provides information on transport- http://bts.ucsf.edu/fdatransportal/ Transporters Homo sapiens TransPortal [17] ers important in the drug discovery process fi cations 3 4 1 Insights into Transporter Classifications: an Outline of Transporters as Drug Targets

including over 750 transporter families, and over 10 000 protein sequences are included in the Transporter Classification Database [5] (URL: www.tcdb.org/). The database stores sequences and information regarding all classified transport proteins. Transporters can be found in the database either by browsing the classification or by directly searching for the protein of interest (e.g., by UniProt accession number). For unclassified proteins, similar sequences can be found using a BLAST search. The classification scheme contains five levels as exemplified for 2.A.22.6.3 in Table 1.2. The first level number indicates the class of membrane transport protein. This can be, for example, a channel or primary active transporter. Interest- ingly, also accessory factors involved in transport are included. Classes 6 and 7 are currently empty and serve as placeholders for yet undiscovered types of transport and class 9 contains not yet fully characterized transporters. Next, a letter indicates the transporter subclass (e.g., energy source for primary active transport), followed by a number for the transporter family or superfamily. The assignment of a transporter to a specific family follows strict statistical criteria of homology, requiring comparison over a region of at least 60 residues and a prob- ability of 10 19 or less than this degree of sequence similarity occurred by coinci- dence [25]. The fourth level indicates the transporter subfamily and the last level classifies a transport system according to its substrate or range of substrates. To summarize, the first two levels describe the function of the transporter, the next two classify according to phylogenetic similarity, and the last one defines the substrate or indicates the belonging to a transport system.

Table 1.2 Transport system 2.A.22.6.3 as an example of the TCDB classiﬁcation.

TCDB level TC number of TCDB name of the exemplary level the level

Transporter class (level 1, 2 Electrochemical potential-driven functional): transporters Transporter subclass (level 2, 2.A Porters (uniporters, symporters, and functional): antiporters) Transporter family/superfamily 2.A.22 The neurotransmitter:sodium symporter (level 3, phylogenetic): family Transporter subfamily (level 4, 2.A.22.6 No explicit level name phylogenetic): TCDB level 5 (examples for the 2.A.22.6.3 No explicit level name, given examples: transport system having the Sodium-dependent neutral amino acid same substrate): transporter B(0)AT1 (human) Sodium-dependent neutral amino acid transporter B(0)AT1 (mouse) Transmembrane protein 27 aka TMEM27 1.2 Available Transporter Classiﬁcations 5

1.2.2 IUPHAR/BPS

The Guide to PHARMACOLOGY database and web page (IUPHAR/BPS Guide to PHARMACOLOGY; URL: www.guidetopharmacology.org/) is created by cooperation between the British Pharmacological Society (BPS) and the Interna- tional Union of Basic and Clinical Pharmacology (IUPHAR). Originally, they provided on their web page access to their two independent databases, the BPS Guide to Receptors and Channels [26] and the IUPHAR database [27]. Since 2014, the IUPHAR database is included in the BPS database and the web page was renamed from BPS Guide to Receptors and Channels to IUPHAR/BPS Guide to PHARMACOLOGY [28]. The IUPHAR/BPS Guide to PHARMACOLOGY is an expert-driven collection of pharmacological targets and the substances that act on them. It contains several different sections, including G protein-coupled receptors (GPCRs), ion channels, nuclear hormone receptors (NHRs), kinases, catalytic receptors, enzymes, other protein targets, and transporters. The transporters are divided into the ATP-binding cassette family, F-type and V-type ATPases, P-Type ATPases, major facilitator superfamily (MFS) of transporters, and the SLC superfamily of solute carriers. The nomenclature follows mainly the HGNC gene families. In addition to the standard SLC nomenclature (SLC family 1–52), some of the SLC families are further divided according to commonality of the substrate. The channels are divided into voltage-gated ion channels, ligand-gated ion channels, and other ion channels.

1.2.3 ChEMBL-16 and ChEMBL-18

The ChEMBL database [6] (URL: https://www.ebi.ac.uk/chembl/) provides large- scale bioactivity data, linking small molecules to the protein targets through which they exert their effects. In order to facilitate browsing and analysis of these data, it was necessary to provide a protein family classification system within the database. The emphasis of the ChEMBL-16 classification was on a functional rather than a sequence-based classification. Since the ChEMBL-16 transporter hierar- chy was heavily focused on protein function, inclusion of new proteins was a largely manual process, relying on the availability of significant knowledge around these proteins. However, the exact transport mechanism (e.g., antiporter, uniporter, Na-symporter, or H-symporter) for a number of transporters, such as the OATPs, is unknown or not completely understood [29]. This makes it diffi- cult to include them in this classification scheme. Therefore, ChEMBL decided, starting with ChEMBL-18 [30], to move to a more phylogenetic classification that is easier to maintain. The classification is derived mainly from IUPHAR/BPS and TCDB. The idea and schema of this combined classification is described below and depicted in Table 1.5. 6 1 Insights into Transporter Classifications: an Outline of Transporters as Drug Targets

1.2.4 SLC Series

In the 1990s, Matthias A. Hediger developed the nomenclature of solute carrier families in collaboration with Phyllis McAlpine [31]. The HUGO Gene Nomen- clature Committee (HGNC, available from www.genenames.org) included this nomenclature for gene names of classical membrane transport proteins. Although originally introduced for human genes, the term is sometimes used for nonhuman species as well. A collection of this 52 family-containing series is available from www. bioparadigms.org. Although the number of included transporters is limited to a specific type of membrane transport proteins, available reviews for each family by experts [12], and manually curated information on transport type, substrates, and expression, make this collection a valuable resource. Members within an individual SLC family share more than 20% sequence similarity with each other, but the homology between the 52 families is often quite low or nonexistent [32]. The SLC members are treated in different ways in other classifications. IUPHAR/BPS summarizes the SLC families in a superfamily. Due to the fact that SLC families have only a vague definition in common (membrane transport proteins that are not driven by ATP) and their sometimes missing sequence similarity, TCDB has no single class that contains all SLC families. SLC families that are not found in class 2 of TCDB (electrochemical potential- driven transporters) are given in Table 1.3.

Table 1.3 SLC members counterparts in TCDB that do not belong to TCDB class 2: electrochemical potential-driven transporters.

SLC family TC class TC family

SLC42 Rh ammonium transporter Channels/pores 1.A.11 the ammonia trans- family porter channel (Amt) family SLC41 MgtE-like magnesium Channels/pores 1.A.26 the Mg2+ transporter- transporter family E (MgtE) family SLC14 urea transporter (UT) Channels/pores 1.A.28 the urea transporter family (UT) family SLC31 copper transporter family Channels/pores 1.A.56 the copper transporter (Ctr) family SLC27 fatty acid transporter Group translocators 4.C.1 the proposed fatty acid family transporter (FAT) family SLC3 Heavy subunits of the het- Accessory factors 8.A.9 the rBAT transport acces- eromeric amino acid transporters involved in transport sory protein (rBAT) family SLC52 riboflavin transporter fam- Incompletely character- 9.A.53 the eukaryotic ribofla- ily RFVT/SLC52 ized transport systems vin transporter (E-RFT) family SLC50 sugar efflux transporters Incompletely character- 9.A.58 the sweet; PQ-loop; ized transport systems saliva; MtN3 (sweet) family 1.4 Merged Top-Level Transporter Classification 7

The SLC proteins belong to several different Pfam [33] clans, thus sharing spe- ciﬁc sequence motifs. The largest ones are the major facilitator superfamily and the amino acid-polyamine-organocation (APC) superfamily, which contain members of 14 and 9 SLC families, respectively [32]. The afﬁliation of the SLC families to a superfamily is depicted in Figure 1.2.

1.3 Function versus Sequence Similarity

Regarding their function, channels and transporters are clearly distinguishable, but protein classifications in our days classify besides functional criteria often according to phylogenetic relationships. Therefore, ABC transporter proteins that act as channels (CFTR (ABCC7), SUR1 (ABCC8), and SUR2 (ABCC9)) or as translation factors (ABCE1, ABCF1, ABCF2, and ABCF3) are assigned as transporters in TCDB due to their sequence similarity to the functional ABC transporters. Also, the majority of proteins in the solute carrier superfamily act as secondary active transporter and, therefore, the 52 SLC families are included as transporters in IUPHAR/BPS. However, for more than a few solute carrier members, the transport mechanism is currently unrevealed and for some it is known that they function as channels. Table 1.4 shows the contradictory classification of some membrane transport proteins in TCDB, IUPHAR/BPS, and ChEMBL-16. Figure 1.1 tries to give an overall impression of human proteins involved in transmembrane transport. Furthermore, basic differences between the more phylogenetic-driven classification TCDB and the more pharmacological-driven classification IUPHAR/BPS can be read out. Figure 1.1 contains some simplification. Not for every IUPHAR/BPS level is an exact counterpart available in TCDB and vice versa. For instance, there is no SLC superfamily in TCDB. Equivalents of SLC families can be found in TCDB class 1, 2, 4, and 8, but they are classified mainly into class 2 (electrochemical potential-driven transporters). Table 1.3 lists the SLC counterparts in TCDB that are not labeled as electrochemical potential-driven transporters. Further- more, the equivalent of an SLC family in TCDB can contain proteins related to this SLC family but not belonging to this SLC family. For instance, the Rh ammonium transporter family (SLC 42) comprises three Rh glycoproteins, namely, RhAG, RhBG, and RhCG. In red blood cells, the ammonium transport is mediated by a complex of RhAG, RhCE, and RhD [34]. The latter two are not included in SLC 42, but due to their function and sequence similarity all five proteins share the same family (1.A.11 the ammonia channel transporter (Amt) family) and subfamily (1.A.11.4) in TCDB.

1.4 Merged Top-Level Transporter Classiﬁcation

For a combined overview on human transporters, Figure 1.2 shows a coarse clas- siﬁcation into four major groups (solute carriers, ATPases, ABC proteins, and 8 1 Insights into Transporter Classiﬁcations: an Outline of Transporters as Drug Targets

Table 1.4 Contradictory classiﬁcation in TCDB, IUPHAR/BPS, and ChEMBL-16.

Protein name Gene name TCDB IUPHAR/BPS ChEMBL-16

Ammonium trans- RHCG C15orf6 Channels/ Transporters — porter Rh type C CDRC2 PDRC2 pores RHGK Ammonium trans- RHBG Channels/ Transporters — porter Rh type B pores Ammonium trans- RHAG RH50 Channels/ Transporters — porter Rh type A pores Solute carrier family 41 SLC41A1 Channels/ Transporters — member 1 pores Solute carrier family 41 SLC41A2 Channels/ Transporters — member 2 pores Solute carrier family 41 SLC41A3 Channels/ Transporters — member 3 pores Urea transporter 1 SLC14A1 HUT11 JK Channels/ Transporters — RACH1 UT1 UTE pores Urea transporter 2 SLC14A2 HUT2 Channels/ Transporters — UT2 pores High-affinity copper SLC31A1 COPT1 Channels/ Transporters — uptake protein 1 CTR1 pores Probable low-affinity SLC31A2 COPT2 Channels/ Transporters — copper uptake protein 2 CTR2 pores ATP-binding cassette ABCE1 RLI RNA- Primary —— subfamily E member 1 SEL1 RNASELI active RNS4I OK/SW-cl.40 transporters ATP-binding cassette ABCF1 ABC50 Primary —— subfamily F member 1 active transporters ATP-binding cassette ABCF3 Primary —— subfamily F member 3 active transporters ATP-binding cassette ABCF2 HUSSY-18 Primary —— subfamily F member 2 active transporters Cystic fibrosis trans- CFTR ABCC7 Primary Ion channels Transporter membrane conduct- active ance regulator transporters ATP-binding cassette ABCC9 SUR2 Primary Transporters Ion channel subfamily C member 9 active transporters ATP-binding cassette ABCC8 HRINS SUR Primary Transporters Ion channel subfamily C member 8 SUR1 active transporters 1.4 Merged Top-Level Transporter Classification 9

Figure 1.1 Overview of human membrane transport proteins. 10 1 Insights into Transporter Classiﬁcations: an Outline of Transporters as Drug Targets

Figure 1.2 Simpliﬁed overview of human transport protein families. Abbreviations: MFS, major facilitator superfamily; APC, amino acid/polyamine/organocation superfamily; CPA, cation:proton antiporter superfamily.

other transporters). Figure 1.2 tries to reflect both human transporter classifications in TCDB and IUPHAR/BPS. The reported names of the protein groups and the number of proteins in Figure 1.2 provide only a rough guide that can vary considerably between the actual classifications. For instance, in a pure functional transporter classification, you may find only 41 human ABC proteins and 381 solute carriers. In a pharmacology-driven classification like IUPHAR/BPS, only the target of levetiracetam (the synaptic vesicle glycoprotein 2A, SV2A) from the group of other transporters is included. Nevertheless, the splitting into four major groups is inspired by the IUPHAR/BPS Guide to PHARMACOLOGY classification of transporters. The assignment to a Pfam clan or family (e.g., the MFS) for the 52 SLC families is color-coded to show their phylogenetic heterogeneity. The MFS group in the group of other transporters is also color-coded, to show the connection to 1.5 Choice and Design of the New ChEMBL Classification 11 proteins that are not assigned to an SLC family (e.g., SV2A). The subgroups of the last group in Figure 1.2 (other transporters) are derived from TCDB.

1.5 Choice and Design of the New ChEMBL Classiﬁcation

Within the framework of the Open PHACTS project [35,36], we were interested to find a classification suitable for channels and transporters. For this, integrating the classification into the existing ChEMBL classification was chosen to facilitate maintainability. By querying different databases (UniProt: reviewed+human+key- word:transport (April 3, 2013); TCDB: all human proteins (May 30, 2013); HGNC: known channel and transporter gene families; GeneOntology: Homo sapiens+GO:0022857 transmembrane transport activity (June 15, 2013)), we compiled a list with 1144 human membrane transport proteins and additionally included 300 nonhuman transporters and channels of ChEMBL-16. The comparison of the classifications for this list of proteins is shown in Figure 1.3. Each of the data sources contains proteins that are unique to this database. Even TCDB, which uses a comprehensive classification of all transport proteins, does not include all identified transporters. An explanation for this is that for

Figure 1.3 Overlap of classiﬁed membrane transport proteins in IUPHAR/BPS, TCDB, and ChEMBL-16. 12 1 Insights into Transporter Classiﬁcations: an Outline of Transporters as Drug Targets

each family, only some examples are provided but not an exhaustive list. On the other hand, TCDB covers even other drug targets such as proteins involved in endocytosis with the term “membrane transport protein.” For instance, targets of the medically used botulinum neurotoxin A, for example, SNAP-25 (TC-ID 1.F.1: the synaptosomal vesicle fusion pore (SVF-Pore) family) [37], and a target of the cholesterol-lowering drug ezetimibe, the Niemann-Pick C-1-like protein (TC-ID 2. A.6: the eukaryotic (putative) sterol transporter family) [38] are included in TCDB. SNAP-25 is not contained in IUPHAR/BPS and Niemann-Pick C-1-like protein is classified into other protein targets and patched family. Nevertheless, IUPHAR/BPS includes the highest number of transport proteins from this list, reflecting its focus on pharmacologically relevant proteins. To generate a classification that can include all proteins from the list, we first predicted, where possible, the classification of the proteins that were unclassified in IUPHAR/BPS (500) or TCDB (604). For proteins where this was not possible in IUPHAR/BPS, new groups were created or added from TCDB. Finally, IUPHAR/BPS was used as the basis for ion channels, including some subclasses of TCDB. For the transporters, a combination of IUPHAR/BPS and TCDB was used, following TCDB for the first and second level, and afterward using an IUPHAR/BPS-based classification (including the concept of an SLC superfamily). This introduced some contradictions, which were accepted as the SLC classification is well known, thus increasing the usability. In addition, a top-level group of auxiliary transport proteins was introduced according to TCDB class 8. Table 1.5 shows human transport protein containing classes and subclasses of TCDB and the equivalent groups in IUPHAR/BPS and ChEMBL-19. Text in italics indicates IUPHAR/BPS as source for the ChEMBL-19 classification.

1.6 Transporter as Drug Targets

To get the first insight into the topic of transporter as drug targets, Table 1.6 shows examples of approved drugs and the targeted transport protein group. The drugs in Table 1.6 are all derived from DrugBank. For some, the exact pharmacological mechanism is largely unknown, for example, artemisinin and derivates. Furthermore, the transporter may be one but not the main target for the indication. For instance, the diuretic effect of amiloride is mainly assigned to the + inhibition of epithelial Na -channels. Table 1.6 also includes some nonclassical transporters (printed in italics). These were included because these may be found in phylogenetic transporter classifications like TCDB and we share with Ashcroft et al. the view of a blurred boundary between channels and transporters [39]. For instance, the cystic fibrosis transmembrane conductance regulator (CFTR) protein acts as a chloride channel but is often classified as ABC transporter due to phylogenetic reasons. Also, the target of ezetimibe, which is nei- ther a functional channel nor a transporter, is included as it was one of the examples given by Imming et al. [1]. Table 1.5 Human membrane transport protein classification of TCDB, IUPHAR/BPS, and ChEMBL-19 in contrast.

TCDB IUPHAR/BPS ChEMBL-19

Channels/pores Ion channels Ion channels

Alpha-type channels Ligand-gated ion channels Ligand-gated ion channels Beta-barrel porins Voltage-gated ion channels Voltage-gated ion channels Pore-forming toxins Other ion channels Other ion channels Vesicle fusion pores Aquaporins Aquaporins Paracellular channels Chloride channels Chloride channels Membrane-bounded channels Connexins and pannexins Connexins and pannexins Sodium-leak channel, nonselective Sodium-leak channel, nonselective Vesicle fusion pores Annexins ... Electrochemical transporter Transporters Transporters

Porters (uniporters, symporters, antiporters) SLC-superfamily of solute carriers Electrochemical transporter Major facilitator superfamily (MFS) of transporters SLC superfamily of solute carriers Primary active transporters ATP-binding cassette transporter family Vesicular neurotransmitter transporter family P-P-bond-hydrolysis-driven transporters P-Type ATPases Primary active transporter Oxidoreduction-driven transporters F-type and V-type ATPases ATP-binding cassette P-Type ATPases Group translocators F-type and V-type ATPases Targets Drug as Transporter 1.6 Acyl CoA ligase-coupled transporters Endoplasmic reticular retrotranslocon family Polysaccharide synthase/exporters Oxidoreduction-driven transporters Group translocator Transmembrane 1-electron transfer carriers Transmembrane electron carriers

Transmembrane 1-electron transfer carriers Accessory factors involved in transport — Auxiliary transport proteins

Auxiliary transport proteins

Incompletely characterized transport systems ——13 14 1 Insights into Transporter Classiﬁcations: an Outline of Transporters as Drug Targets

Table 1.6 Approved drugs and targeted transport proteins.

Major Classiﬁcation in IUPHAR/BPS Example for Drug group group (TC family and TC-ID of the approved drug (ATC code) targeted protein)

Solute SLC2 facilitative GLUT transporter Dapaglifocin Antidiabetic drug carrier family (solute:sodium symporter (A10BX09) (SSS) family; 2.A.21.3.16) SLC6 sodium- and chloride-depen- Fluoxetin Antidepressant dent neurotransmitter transporter (N06AB03) family (neurotransmitter:sodium symporter family; 2.A.22.1.1) SLC6 sodium- and chloride-depen- Tiagabine Antiepileptics dent neurotransmitter transporter (N03AG06) family (neurotransmitter:sodium symporter family; 2.A.22.3.2) SLC9 Na+/H+ exchanger family (the Amiloride Diuretics, potas- monovalent cation:proton antiporter- sium-sparing 1 (CPA1) family; 2.A.36.1.13) (C03DB01) SLC12 electroneutral cation-coupled Furosemide Diuretics, high- Cl cotransporter family (cation-chlo- ceiling (C03CA01) ride cotransporter family; 2.A.30.1.2) SLC18 vesicular amine transporter Reserpine Antihypertensives family (the drug:H+ antiporter-1 (12 (C02AA02) spanner) (DHA1) family; 2.A.1.2.29) Tetrabenazine Hyperkinetic movement dis- order (N07XX06) SLC22 organic cation/anion/zwitter- Probenecid Uricosuric drug ion transporter family (organic cation (M04AB01) transporter (OCT) family; 2. A.1.19.10|2.A.1.19.31|2.A.1.19.34) SLC25 mitochondrial carrier (MC) Clodronate Osteoporosis, family (MC family; 2.A.29.1.2|2. bone metastases A.29.1.1|2.A.29.1.10) (M05BA02) SLC52 riboﬂavin transporter family Gamma hydroxy- Anesthetics RFVT/SLC52 (E-RFT family; 9. butyric acid (N01AX11) A.53.1.3) ATPases P-type ATPase (P-type ATPase Digitoxin Cardiac glycosides superfamily; 3.A.3.1.1) (c1AA04) P-type ATPase (P-type ATPase Omeprazol Proton pump superfamily; 3.A.3.1.2) inhibitors (A02Bc1) P-type ATPase (P-type ATPase Lumefantrine, Antimalarials superfamily; 3.A.3.2.?) artemether (arte- (P01BF01) misinin derivates) 1.7 Drug Targets in the SLC Classiﬁcation 15

Table 1.6 (continued ) Major Classiﬁcation in IUPHAR/BPS Example for Drug group group (TC family and TC-ID of the approved drug (ATC code) targeted protein)

ABC ATP-binding cassette subfamily C Ivacaftor Cystic ﬁbrosis protein member 7 ( cystic ﬁbrosis trans- (R07AX02) membrane conductance exporter (CFTR) family; 3.A.1.202.1) ATP-binding cassette subfamily C Repaglinide Antidiabetic drug member 8 (the drug conjugate trans- (A10BX02) porter (DCT) family; 3.A.1.208.4) Other Major facilitator superfamily of Levetiracetam Antiepileptics transport transporters, non-SLC (vesicular (N03AX14) proteins neurotransmitter transporter (VNT) family; 2.A.1.22.?) Other protein targets (eukaryotic Ezetimibe Lipid-modifying (putative) sterol transporter ( EST) agent (C10AX09) family; 2.A.6.6.6)

Note: The transport proteins are not necessarily the target responsible for the reported drug indication.

Regarding prospective targets, various transporters are and were considered promising drug targets in chemotherapy, but so far the candidates have failed in clinical trials. At present, Winter et al. describe SLC35F2 as a prospective target in cancer therapy and Rask-Andersen et al. mention members of SLC2, SLC5, SLC7, and SLC9 as promising targets in cancer therapy and members of SLC10 as potential targets against constipation and hypercholesterolemia [40,41].

1.7 Drug Targets in the SLC Classiﬁcation

Several solute carriers are reported targets of approved drugs [41]. Here, we use the SLC classification as a framework to give an overview on the diseases a transporter is connected to, known drug molecules that directly target the transporter, and a count of bioactivity data available in ChEMBL to give an estimate on the degree of interest in the target. The data shown in the figures was collected from different sources. Molecules targeting transporters were retrieved from the DrugBank xml [42], using a modi- fied KNIME [43] workflow from the available example workflows [44]. Disease information was downloaded from DisGeNET [45] using the curated gene– disease associations [46]. Protein/gene mappings were generated from UniProt [23]. Bioactivity data counts were retrieved from ChEMBL-19, using single proteins only. Cladograms were generated with FigTree [47] using SLC sequences retrieved from Uniprot. Multiple sequence alignments for the 16 1 Insights into Transporter Classifications: an Outline of Transporters as Drug Targets

members of one Pfam clan (e.g., the major facilitator superfamily) were generated with Clustal Omega [48] using the default parameters on the EBI Web server [49]. Counts for each target were added manually. Figure 1.4 shows the counts for SLC members belonging to the amino acid- polyamine-organocation superfamily. Seven of the families show members with reported drugs, with three of them being previously reported by Rask-Andersen

Figure 1.4 Cladogram of the 9 SLC families belonging to the APC superfamily and the number of connected diseases, drugs, and bioactivity values. 1.8 Conclusions 17 et al. [41] to be targets of approved drugs or under investigation (SLC5, SLC7, and SLC12). Closer investigation of the drugs for the remaining families shows that these are mostly vitamins or amino acids. Investigating the number of associated diseases for families without known drugs ﬁnds SLC4 and SLC26 as interesting families. Indeed, these are mentioned as potential new targets by Rask-Andersen et al., however, as target for antineoplastic agents, which is not one of the associated diseases. Figure 1.4 thus shows their association with several other diseases as well.

1.8 Conclusions

In this chapter, we gave an overview of available transporter classification schemes. A new version of the ChEMBL classification was introduced. For this, we wanted to have a less complex, browsable classification and, therefore, merged TCDB with the IUPHAR/BPS classification. The advantage compared to the pure IUPHAR/BPS transporter classification is that you still easily find the main transporter groups (ABC transporter, SLC members, and ATPases) and, if new bioactivity data for less common human or nonhuman transporters/transporter families are reported, these transporters can be easily integrated in con- formity with TCDB, which is more complicated with a classification following IUPHAR/BPS only. For ChEMBL, we wanted to use the well-known SLC families to have a less complex transport protein classification than TCDB but keep the possibility to extend the scheme with the corresponding TCDB classes if it becomes significant. Classifications allow (semi)automatic clustering of information. We used the SLC families to give an overview of interacting drugs and associated diseases of members of the APC clan. One disadvantage of an automated approach, however, is that false positive connections can be drawn. For example, the only human member of SLC32, the vesicular inhibitory amino acid transporter (VIAAT), seems to have a target drug according to Figure 1.3. On closer inspec- tion, this is glycine, which is one of the natural substrates of this transporter. A more detailed investigation will, therefore, be necessary to draw valid conclusions from these investigations.

Acknowledgment

The research leading to these results has received support from the Innovative Medicines Initiative Joint Undertaking under grant agreement no. 115191, resources of which are composed of financial contribution from the European Union’s Seventh Framework Programme (FP7/2007-2013) and EFPIA compa- nies’ in-kind contribution. We also acknowledge financial support provided by the Austrian Science Fund, grant F3502. 18 1 Insights into Transporter Classifications: an Outline of Transporters as Drug Targets

References

1 Imming, P., Sinning, C., and Meyer, A. multidrug resistance efﬂux pumps of the (2006) Drugs, their targets and the nature major facilitator superfamily. Int. J. and number of drug targets. Nat. Rev. Bacteriol., 2013, e204141. Drug Discov., 5 (10), 821–834. 11 Slavic, K., Krishna, S., Derbyshire, E.T., 2 Rask-Andersen, M., Almén, M.S., and and Staines, H.M. (2011) Plasmodial sugar Schiöth, H.B. (2011) Trends in the transporters as anti-malarial drug targets exploitation of novel drug targets. Nat. and comparisons with other protozoa. Rev. Drug Discov., 10 (8), 579–590. Malar. J., 10 (1), 165. 3 Wishart, D.S., Knox, C., Guo, A.C., Cheng, 12 Hediger, M.A., Clémençon, B., Burrier, D., Shrivastava, S., Tzur, D., Gautam, B., R.E., and Bruford, E.A. (2013) The and Hassanali, M. (2008) DrugBank: a ABCs of membrane transporters in knowledgebase for drugs, drug actions and health and disease (SLC series): drug targets. Nucleic Acids Res., 36 introduction. Mol. Aspects Med., (Database issue), D901–D906. 34 (2–3), 95–107. 4 Rask-Andersen, M., Masuram, S., and 13 M. Müller. ABC-Transporter Proteins Schiöth, H.B. (2014) The druggable and other Transporters. Available at genome: evaluation of drug targets in http://nutrigene.4t.com/translink.htm clinical trials suggests major shifts in (accessed 22 Sep 2016). molecular class and indication. Annu. 14 Pawson, A.J., Sharman, J.L., Benson, H.E., Rev. Pharmacol. Toxicol., 54 (1), Faccenda, E., Alexander, S.P.H., Buneman, 9–26. O.P., Davenport, A.P., McGrath, J.C., 5 Saier, M.H., Reddy, V.S., Tamang, D.G., Peters, J.A., Southan, C., Spedding, M., Yu, and Vastermark, A. (2013) The transporter W., and Harmar, A.J. (2014) The classiﬁcation database. Nucleic Acids Res., IUPHAR/BPS Guide to 42 (D1), D251–D258. PHARMACOLOGY: an expert-driven 6 Gaulton, A., Bellis, L.J., Bento, A.P., knowledgebase of drug targets and their Chambers, J., Davies, M., Hersey, A., Light, ligands. Nucleic Acids Res., 42 (Database Y., McGlinchey, S., Michalovich, D., Al- issue), D1098–D1106. Lazikani, B., and Overington, J.P. (2011) 15 Ren, Q., Chen, K., and Paulsen, I.T. (2007) ChEMBL: a large-scale bioactivity database TransportDB: a comprehensive database for drug discovery. Nucleic Acids Res., resource for cytoplasmic membrane 40 (D1), D1100–D1107. transport systems and outer membrane 7 Schwacke, R., Schneider, A., van der channels. Nucleic Acids Res., Graaff, E., Fischer, K., Catoni, E., 35 (Database), D274–D279. Desimone, M., Frommer, W.B., Flügge, U.- 16 Zhao, M., Chen, Y., Qu, D., and Qu, H. I., and Kunze, R. (2003) ARAMEMNON, a (2011) TSdb: a database of transporter novel database for arabidopsis integral substrates linking metabolic pathways and membrane proteins. Plant Physiol., transporter systems on a genome scale via 131 (1), 16–26. their shared substrates. Sci. China Life Sci., 8 Brohée, S., Barriot, R., Moreau, Y., and 54 (1), 60–64. André, B. (2010) YTPdb: a wiki database 17 Morrissey, K.M., Wen, C.C., Johns, S.J., of yeast membrane transporters. Biochim. Zhang, L., Huang, S.-M., and Giacomini, Biophys. Acta, 1798 (10), 1908–1912. K.M. (2012) The UCSF-FDA TransPortal: 9 Fichant, G., Basse, M.-J., and Quentin, Y. a public drug transporter database. (2006) ABCdb: an online resource for Clin. Pharmacol. Ther., 92 (5), ABC transporter repertories from 545–546. sequenced archaeal and bacterial 18 Hersey, A. (Sep. 2013) ChEMBL database genomes. FEMS Microbiol. Lett., release 16, EMBL-EBI. 256 (2), 333–339. 19 Viereck, M., Gaulton, A., Digles, D., and 10 Kumar, S., Mukherjee, M.M., and Varela, Ecker, G.F. (2014) Transporter taxonomy: M.F. (2013) Modulation of bacterial a comparison of different transport References 19

protein classification schemes. Drug implications of human membrane Discov. Today Technol., 12, e37–e46. transport proteins. Pflügers Arch., 447 (5), 20 Saier, M.H. (2000) A functional- 465–468. phylogenetic classification system for 32 Höglund, P.J., Nordström, K.J.V., Schiöth, transmembrane solute transporters. H.B., and Fredriksson, R. (2011) The Microbiol. Mol. Biol. Rev., 64 (2), 354–411. solute carrier families have a remarkably 21 Tipton, K. and Boyce, S. (2000) History of long evolutionary history with the majority the enzyme nomenclature system. of the human families present before Bioinforma. Oxf. Engl., 16 (1), 34–40. divergence of bilaterian species. Mol. Biol. 22 Busch, W. and Saier, M.H., Jr. (2003) The Evol., 28 (4), 1531–1541. IUBMB-endorsed transporter classification 33 Punta, M., Coggill, P.C., Eberhardt, R.Y., system. Methods Mol. Biol., 227,21–36. Mistry, J., Tate, J., Boursnell, C., Pang, N., 23 The UniProt Consortium (2013) Activities Forslund, K., Ceric, G., Clements, J., at the universal protein resource (UniProt). Heger, A., Holm, L., Sonnhammer, E.L.L., Nucleic Acids Res., 42 (D1), D191–D198. Eddy, S.R., Bateman, A., and Finn, R.D. 24 Berman, H.M., Westbrook, J., Feng, Z., (2011) The Pfam protein families database. Gilliland, G., Bhat, T.N., Weissig, H., Nucleic Acids Res., 40 (D1), D290–D301. Shindyalov, I.N., and Bourne, P.E. (2000) 34 Nakhoul, N.L. and Lee Hamm, L. (2013) The protein data bank. Nucleic Acids Res., Characteristics of mammalian Rh 28 (1), 235–242. glycoproteins (SLC42 transporters) and 25 Saier, M.H., Yen, M.R., Noto, K., Tamang, their role in acid–base transport. Mol. D.G., and Elkan, C. (2009) The transporter Aspects Med., 34 (2–3), 629–637. classification database: recent advances. 35 Gray, A.J.G., Groth, P., Loizou, A., Askjaer, Nucleic Acids Res., 37 (Database), S., Brenninkmeijer, C., Burger, K., D274–D278. Chichester, C., Evelo, C.T., Goble, C., 26 Alexander, S., Mathie, A., and Peters, J. Harland, L., Pettifer, S., Thompson, M., (2011) Guide to receptors and channels Waagmeester, A., and Williams, A.J. (GRAC). 5th edition, Br. J. Pharmacol., (2014) Applying linked data approaches 164,S1–S2. to pharmacology: architectural decisions 27 Sharman, J.L., Benson, H.E., Pawson, A.J., and implementation. Semantic Web, 5, Lukito, V., Mpamhanga, C.P., Bombail, V., 101–113. Davenport, A.P., Peters, J.A., Spedding, M., 36 Williams, A.J., Harland, L., Groth, P., Harmar, A.J., and NC-IUPHAR (2013) Pettifer, S., Chichester, C., Willighagen, IUPHAR-DB: updated database content E.L., Evelo, C.T., Blomberg, N., Ecker, G., and new features. Nucleic Acids Res., Goble, C., and Mons, B. (2012) Open 41 (D1), D1083–D1088. PHACTS: semantic interoperability for 28 Alexander, S.P.H., Benson, H.E., Faccenda, drug discovery. Drug Discov. Today, E., Pawson, A.J., Sharman, J.L., McGrath, 17 (21–22), 1188–1198. J.C., Catterall, W.A., Spedding, M., Peters, 37 Frevert, J. (2015) Pharmaceutical, biological, J.A., Harmar, A.J., and CGTP and clinical properties of botulinum Collaborators (2013) The Concise Guide neurotoxintypeaproducts.Drugs RD, 15, to PHARMACOLOGY 2013/14: overview. 1–9. Br. J. Pharmacol., 170 (8), 1449–1458. 38 Altmann, S.W., Davis, H.R., Zhu, L., Yao, 29 Hagenbuch, B. and Stieger, B. (2013) The X., Hoos, L.M., Tetzloff, G., Iyer, S.P.N., SLCO (former SLC21) superfamily of Maguire, M., Golovko, A., Zeng, M., transporters. Mol. Aspects Med., 34 (2–3), Wang, L., Murgolo, N., and Graziano, 396–412. M.P. (2004) Niemann-Pick C1 like 1 30 Hersey, A. (Aug. (2014)) CHEMBL protein is critical for intestinal cholesterol database release 18, EMBL-EBI. absorption. Science, 303 (5661), 31 Hediger, M.A., Romero, M.F., Peng, J.-B., 1201–1204. Rolfs, A., Takanaga, H., and Bruford, E.A. 39 Ashcroft, F., Gadsby, D., and Miller, C. (2004) The ABCs of solute carriers: (2009) Introduction. The blurred physiological, pathological and therapeutic boundary between channels and 20 1 Insights into Transporter Classifications: an Outline of Transporters as Drug Targets

transporters. Philos. Trans. R. Soc. B, 45 Bauer-Mehren, A., Bundschus, M., 364 (1514), 145–147. Rautschka, M., Mayer, M.A., Sanz, F., 40 Winter, G.E., Radic, B., Mayor-Ruiz, C., andFurlong,L.I.(2011)Gene-disease Blomen, V.A., Trefzer, C., Kandasamy, network analysis reveals functional R.K., Huber, K.V.M., Gridling, M., Chen, modules in mendelian, complex and D., Klampfl, T., Kralovics, R., Kubicek, S., environmental diseases. PLoS ONE, Fernandez-Capetillo, O., Brummelkamp, 6 (6), e20284. T.R., and Superti-Furga, G. (2014) The 46 DisGeNET: a database of gene-disease solute carrier SLC35F2 enables YM155- associations,” DisGeNET. Available at mediated DNA damage toxicity. Nat. http://www.disgenet.org/web/DisGeNET/ Chem. Biol., 10 (9), 768–773. v2.1/downloads#curated (accessed 2 41 Rask-Andersen, M., Masuram, S., October 2014). Fredriksson, R., and Schiöth, H.B. (2013) 47 FigTree v1.4.2,” Molecular Evolution, Solute carriers as drug targets: current use, Phylogenetics and Epidemiology. Available clinical trials and prospective. Mol. Aspects at http://tree.bio.ed.ac.uk/software/figtree/ Med., 34 (2–3), 702–710. (accessed 11 November 2014). 42 DrugBank: Downloads,” DrugBank. 48 Sievers, F., Wilm, A., Dineen, D., Gibson, Available at http://www.drugbank.ca/ T.J., Karplus, K., Li, W., Lopez, R., system/downloads/current/drugbank.xml. McWilliam, H., Remmert, M., Söding, J., zip (accessed 10 October 2014). Thompson, J.D., and Higgins, D.G. (2011) 43 Berthold, M.R., Cebron, N., Dill, F., Fast, scalable generation of high-quality Gabriel, T.R., Kötter, T., Meinl, T., Ohl, P., protein multiple sequence alignments Sieb, C., Thiel, K., and Wiswedel, B. (2007) using Clustal Omega. Mol. Syst. Biol., KNIME: the Konstanz information miner, 7 (1), 539. in Studies in Classification, Data Analysis, 49 Clustal Omega < Multiple Sequence and Knowledge Organization Springer. Alignment < EMBL-EBI.” Available at 44 KNIME | DrugBank,” KNIME. Available at http://www.ebi.ac.uk/Tools/msa/clustalo/ http://tech.knime.org/book/network- (accessed: 12 October 2014). examples/drugBank (accessed 1 July 2014).